Recent scientific and clinical interest in combining therapeutic agents for the treatment of alcoholism are based on the fact that derangement of multiple-neurotransmitter systems are likely to underlie biological predisposition to the disease. Thus, combining effective medications working at different neurotransmitters should produce a synergistic or at least an added clinical response. In animals, the combination of the 5-HT3 antagonist, ondansetron, and the mu receptor antagonist, naltrexone show synergism of action at reducing ethanol consumption. Alcoholics with an early onset of disease are effectively treated by ondansetron, and those with a family history of alcoholism in first degree relatives may have the best clinical outcome to treatment with naltrexone. Given that family history of alcoholism is associated with an early onset of disease, it reasonable for us to predict that the combination of ondansetron and naltrexone should be more optimal than either alone for the treatment of Early Onset Alcoholics (EOA). Indeed, preliminary clinical data from our group provide strong support that the medication combination is an effective treatment for EOA. We will test this hypothesis by comparing the effectiveness of ondansetron (4 mg/kg) and naltrexone(50 mg/day), both alone and in combination, in treating EOA vs. Late Onset Alcoholics (LOA) (total N of 45 subjects/cell x 8 cells = 360) in a randomized, double-blind, placebo-controlled, 12-week (1 week of single-blind placebo followed by 11 weeks of the double-blind condition) outpatient clinical trial. All subjects will receive standardized Cognitive Behavioral Therapy, and follow-up at 1, 3, 6, and 9 months post-treatment. Specifically, we predict that: 1) EOA, compared with LOA, will be more responsive to treatment with either ondansetron or naltrexone alone, and 2) that the combination of ondansetron and naltrexone will be superior to either medication alone in the treatment of EOA. We will have the unique opportunity to test with adequate power the secondary hypothesis that the combination of ondansetron and naltrexone will be better tolerated than naltrexone alone, thereby improving compliance. This is because nausea is an important side-effect of naltrexone which can limit compliance, and as shown in our preliminary study, ondansetron by having anti-nausea and anti-emetic properties counteracts this naltrexone side-effect. We support NIAAA's mission to develop effective pharmacotherapies as adjuncts to psychotherapy for the treatment of alcoholism.